Send to

Choose Destination
Dev Med Child Neurol. 2015 Apr;57(4):358-65. doi: 10.1111/dmcn.12616. Epub 2014 Nov 7.

Validity and reproducibility of measures of oropharyngeal dysphagia in preschool children with cerebral palsy.

Author information

Queensland Cerebral Palsy and Rehabilitation Research Centre, The University of Queensland, Brisbane, Qld, Australia.



The aim of the study was to determine the best measure to discriminate between those with oropharyngeal dysphagia (OPD) and those without OPD, among young children with cerebral palsy (CP).


We carried out a cross-sectional population-based study involving 130 children with CP aged between 18 months and 36 months (mean 27.4mo; 81 males, 49 females) classified according to the Gross Motor Function Classification Scale (GMFCS) as level I (n=57), II (n=15), III (n=23), IV (n=12), or V (n=23). Forty children with CP (mean 28.5mo; 21 males,19 females, eight for each GMFCS level) were included in the reproducibility sub-study, and 40 children with typical development (mean 26.2mo; 18 males, 22 females) were included in the validity sub-study. OPD was assessed using the Dysphagia Disorders Survey (DDS), Pre-Speech Assessment Scale (PSAS), and Schedule for Oral Motor Assessment (SOMA). We analysed reproducibility using inter- and intrarater agreement (percentage) and reliability (kappa values and intraclass correlation coefficients). Construct validity was assessed as concordance between measures (SOMA, DDS, and PSAS). In the absence of a criterion standard measure for OPD, prevalence was estimated using latent class variable analysis. Data from the children with typical development were used to propose modified OPD cut-points for discriminative validity.


All measures had strong agreement (>85%) for inter- and intrarater reliability. The SOMA had the best specificity (100.0%), but lacked sensitivity (53.0%), whereas the DDS and PSAS had high sensitivity (each 100.0%) but lacked specificity (47.1% and 70.6% respectively). OPD prevalence when calculated using the web-based estimation was 65.4%, which was similar to the estimate from the modified cut-points.


Using the sample of children with typical development and modified cut-points, OPD prevalence was lower than estimates with standard scoring. We propose using these modified cut-points when administering the DDS, PSAS or SOMA in young children with CP.

[Indexed for MEDLINE]
Free full text

Supplemental Content

Full text links

Icon for Wiley
Loading ...
Support Center